i

Carbon isotope variations on ancient carbonate platforms: The roles of organic carbon burial and sea level

Honors Research Thesis

Presented in Partial Fulfillment of the Requirements for graduation

“with Honors Research Distinction in Earth Sciences”

in the undergraduate colleges of The Ohio State University

By Cody R. Trigg

The Ohio State University

Spring, 2013

ii

Table of Contents

Abstract…………………………………………………………………………………………………………..i

Introduction…………………………………………………………………………………………………….1

Geologic Setting……………………………………………………………………………………………….2

Methods………………………………………………………………………………………………………….3

Results……………………………………………………………………………………………………………4

Discussion………………………………………………………………………………………………………4

Conclusions and Implications…………………………………………………………………………..8

Recommendations for Future Work…………………………………………………………………..8

Acknowledgements………………………………………………………………………………………….9

References………………………………………………………………………………………………………9

i

Abstract

Positive 13C trends in ancient carbonate rocks are commonly interpreted to

reflect increases in the burial of organic carbon in sediments. An alternative model to

explain positive 13C trends in shelf carbonate rocks is related to mixing with open ocean

water during transgression that floods the carbonate platform. Rock samples from two

age-correlated, stratigraphic sections from the Arbuckle Mountains, Oklahoma and

Clear Spring, Maryland of Middle to Upper Ordovician age were measured and sampled

for 13C and 18O. The 13C curves reveal 3-4 per mil positive shifts in the upper

Darriwilian to lower Sandbian stages (C. sweeti to gerdae conodont zones). Lithological

analysis indicates transgressive events (shallow to deeper marine) associated with the

positive 13C shifts in the Appalachian foreland basin but indicates a transgressive-

regressive sequence in the upper portions of the Arbuckle Mountains. Plots of 13C

versus 18O reveal minimal covariance, potentially invalidating diagenetic explanations

for the trend. Because 13C does not show an obvious correlation with water depth or

diagenesis, these results suggest that the positive 13C positive trends of both sections

may be closely related to global changes in organic carbon burial.

1

Introduction

Geochemical properties of carbonate rocks have long been utilized to reconstruct

various aspects of ancient ocean chemistry and climate. In particular, isotopic

indicators, such as 13C and 18O from carbonate, are commonly interpreted as proxies

for changes in carbon dioxide and temperature, respectively, throughout geologic time.

The interpretations for these isotopic indicators hinge on the assumption that the

isotopic composition of the carbonate being measured reflects that of the average, global

surface ocean. This implies that the isotopic signal must be both primary (unaltered)

and representative of contemporaneous, well-mixed ocean water. While quantitative

studies have been done to demonstrate pathways in which carbonates and, specifically,

13C are susceptible to diagenetic alteration (Banner and Hanson, 1990; Allan and

Matthews, 1982), few have focused on the processes which can result in the 13C values

of ancient carbonate platforms to differ from the δ13C composition of the ancient global

ocean system (Immenhauser et al., 2002). These latter processes, which are

oceanographic effects, are related to the convolution of geochemically distinct water

masses of the open ocean and restricted epeiric seas.

From studies of the modern ocean it is known that the 13C of seawater dissolved

inorganic carbon (DIC) can vary significantly with oceanographic setting, with modern

carbonate platforms being up to 4 per mil lighter than that of open ocean water

(Patterson and Walter 1994). In general, it is the restricted circulation and long

residence time of water in the platform setting which allows for the accumulation of 12C-

enriched carbon dioxide from oxidized organic matter. Although these oceanographic

effects are expressed as lateral variations in 13C, it has been postulated that they can be

2

recorded in stratigraphic profiles during the rise or fall of local sea-level (e.g. mixing of

open ocean and restricted platform water masses, Holmden et al., 1998; Immenhauser

et al., 2002; Immenhauser et al., 2003; Fanton and Holmden, 2007). Whereas positive

shifts in 13C are often interpreted to reflect periods of increased burial of organic matter

in sediments (global carbon cycle changes), the above references attempt to show that

local and regional effects associated with sea-level changes can be the dominate

processes affecting 13C. In order to test the hypothesis that 13C curves from carbonates

deposited in epeiric seas record local oceanographic effects, I examined the

relationships among 13C, 18O, and sea-level from two Ordovician stratigraphic sections

during intervals with known transgressive-regressive sequences.

Geologic Setting

Sections for this study were sampled from Middle-Late Ordovician strata along I-

35 in the Arbuckle Mountains of Oklahoma and along I-70 in the Appalachian

Mountains near Clear Spring, Maryland. The Ordovician strata of the Arbuckle

Mountains were deposited in a widespread epeiric seaway and sat adjacent to the

Southern Oklahoma Aulacogen (Bambach et al., 1980) with facies ranging from the

inner ramp to the outer ramp (Carlucci et al., 2012). The St. Paul Group and

Chambersburg Formation of Maryland were deposited on a carbonate platform

associated with the flexurally subsiding Appalachian foreland basin system, with the

base of the St. Paul Group marking the Sauk-Tippecanoe megasequence boundary

(Read, 1980; Brezinski et al., 2012).

3

Methods

The two sections utilized of the Middle to Upper Ordovician are correlated based

on conodont index fossils (C. sweeti to gerdae zones) and appear to exhibit a

contemporaneous, positive 13C trend. For 13C and 18O analyses, rock specimens were

cut with a water-based, diamond-blade saw and sonicated in a water bath to expose

fresh, unweathered surfaces. Fresh surfaces were microdrilled to produce a fine powder

of matrix micrite with special care given to avoiding skeletal components and calcitic

veins. A 75-95µg carbonate subsample was processed for each specimen in a Finnigan

Delta Plus IV coupled with a Kiel III carbonate device stable isotope mass spectrometer,

with standard deviations of repeated measurements from an internal standard of ±

0.03‰. Approximately 10% of all samples were run in duplicates with no significant

variations. Although brachiopod shells are often considered the most reliable form of

carbonate to conduct δ13C analysis because of their resistance to diagenesis, these shells

are not always sufficiently abundant to construct secular 13C trends. Matrix micrite is

thus chosen because of its abundance and in order to test its robustness during changes

in local sea-level, given that the use of matrix micrite for 13C analysis has become

widespread (Saltzman et al., 1998; Payne et al., 2004; Saltzman and Young, 2005;

Johnston et al. 2012). The sequence stratigraphic framework for this study of the

Bromide Formation is from Carlucci et al. (2012), who worked at the same outcrop

which was sampled for this study. For the St. Paul Group and Chambersburg Formation,

sequence stratigraphy was also based on interpretations of lithostratigraphic

relationships but within a more regional context by Brezinski et al. (2012).

4

Results

Carbon isotope curves from Ordovician strata of the Arbuckle Mountains,

Oklahoma and Clear Spring, Maryland both reveal a long-term positive shift in 13C of

approximately 3‰ magnitude beginning in the upper Darriwilian and continuing

through the Sandbian. Sequence stratigraphy from the Bromide Formation in the

Arbuckle Mountains indicates three third-order depositional sequences superimposed

on a second-order, transgressive-regressive sequence (Carlucci et al., 2012). 13C from

the Bromide exhibits a gradual, positive trend beginning at -3‰ and ending near 0‰

and is punctuated with two short negative shifts of ~1‰ magnitude (Fig. 2).

The St. Paul group and the Chambersburg Formation from Clear Spring display a

third order, regressive-transgressive sequence, with transgression occurring just before

the base of the Chambersburg Formation. 13C begins between 0‰ and 1‰, decreases

to just above -4‰ near the end of the Darriwilian, and increases to just below 1‰ into

the Mohawkian where the positive trend appears to continue. The increase in 13C from

its minimum of -4‰ precedes the onset of transgression (Fig. 3). Cross plots of 13C

versus 18O indicate no significant correlation at either locality (Fig. 4).

Discussion

Results of δ18O from these sections reveals effects from diagenesis based on the

generally scattered or invariant trends seen in both sections, as is often the case for

bulk-rock sampling for δ18O analysis. Consequently, δ18O in bulk carbonate has little use

for reconstructing temperature or paleoceanographic trends. However, because cross

plots of δ13C versus δ18O reveal no correlation (Fig. 4), it appears that meteoric

5

diagenetic overprinting was not a significant process which altered δ13C along with δ18O.

As a result, the focus of this discussion will be on δ13C and its relationship to sea-level.

In order to assess the significance of local sea-level and water mass residence

time on the carbon isotope record, I compared our stratigraphic trends of δ13C with the

interpreted sequence stratigraphy at both localities. The strata from the Arbuckle

Mountains display a general positive trend in δ13C with two punctuated negative δ13C

shifts. As seen in Fig. 2, the first portion of the section begins with a positive trend in

δ13C and is contained within a transgressive and a highstrand systems tract. This trend

reverses at the first sequence boundary with a negative shift in δ13C which coincides with

another transgressive event. The negative shift continues into the highstand portion of

the sequence where it inflects into a positive trend and continues as such through falling

stand, lowstand, transgressive, and highstand systems tracts. From the St. Paul Group

and Chambersburg formation of Clear Spring, there is a negative shift associated with

regression, but an inflection toward a positive shift which precedes the onset of

transgression (Fig. 3).

The overall relationship observed here between sequence stratigraphy and δ13C at

both localities suggest that sea-level fluctuations are not the dominant mechanism

governing changes in δ13C. In contrast to the ocean-mixing model proposed by

Immenhauser et al. (2003), I do not consistently observe positive shifts associated with

transgression or negative shifts associated with regression or lowstand tracts (Fig. 5).

However, worth noting is that the negative shift seen in the Upper Darriwilian from

Clear Spring is of greater magnitude than is documented elsewhere (Saltzman and

Young, 2005; Ainsaar et al., 2010; Schmitz et al., 2010). I suggest the large magnitude

can be attributed to seawater ‘ageing’ processes being superimposed on a global,

6

negative shift in δ13C. This ‘ageing’ process can occur during a period of restricted

circulation which enables the accumulation CO2 from oxidized organic matter enriched

in 12C (Holmden et al., 1998). This CO2 will form bicarbonate ions where it can be

incorporated into the skeletal material of carbonate-forming organisms and record the

δ13C of the locally altered sea-water.

The correspondence between the positive shift in δ13C seen in Oklahoma,

Maryland (this study) and in the Antelope Range, Nevada (Saltzman and Young, 2005)

which are each associated with very different tectonic and regional settings implies that

the shift is being driven by changes in the global carbon cycle as a result of increased

rates of organic carbon burial. Although Patterson and Walter (1994) show that modern

restricted platforms can become geochemically distinct from the open ocean due to local

and regional processes, the difference in δ13C largely depends on the proximity to the

open ocean, with extreme cases like Florida Bay being close to a large volume of

freshwater where inputs of weathered carbon are high (Fanton and Holmden, 2007). It

is not likely that the carbonates from these sections were deposited in such extreme

cases of restricted circulation, similar to that of the modern Florida Bay due to the

presence of conodont microfossils throughout our sections. More likely, is a type of

setting similar to the Bahama Banks where there is some, but limited exchange with the

open ocean and the δ13C of ΣCO2 can often be within 1‰ of that of open ocean water

(see Patterson and Walter, 1994). The conclusions derived from Immenhauser and

others (2002, 2003) which suggest a close coupling between δ13C and sea-level are

based on isotopic data from sections with limited controls for stratigraphic correlation

(e.g. marker beds). As such, their study may potentially compare units of different time

or with hiatuses which may affect the interpretations. If the observed positive shift seen

7

by Immenhauser and others (2002, 2003) is an artifact of switching to an open marine

water mass with a different isotopic composition, one would not expect the trend to

reverse unless the oceanographic setting also reversed due to regression and eventually

a lowstand systems tract. Upon close inspection of the sequence stratigraphy following

the positive shift in δ13C, it is not clear that there was a regressive event which could

have caused the negative shift which follows the positive shift. This suggests that the

shift may be related to changes in organic carbon burial.

Furthermore, if the interval studied by Immenhauser and others (2002, 2003)

from the Pennsylvanian occurred during a period of favorable aragonite production,

their observed shifts in δ13C may be a result of an increased presence of 13C-enriched

aragonite sediments associated with sea-level rise, as described by Swart (2008). During

the dominantly calcitic seas of the early Paleozoic (see Dickson, 2002 and references

therein for a discussion of calcite-aragonite seas), it is unlikely that a shift to aragonite

would occur and produce the major variations in δ13C that are recorded in the

Ordovician sections from this study. Additionally, in contrast to the findings of Fanton

and Holmden (2007) which suggest that positive changes in δ13C may be related to

large-scale episodes of continental flooding, and, specifically, that sea-level changes

result in δ13C changes due to local C-cycle perturbations, I observe an overall positive

trend in the Arbuckle Mountains during a high-order transgressive-regressive sequence

in which a local increase in primary productivity is unlikely to account for the shift.

Further, although Fanton and Holmden (2007) suggest their record of δ13C during the

Middle to Late Ordovician is closely related to changes in sea-level, the relationship

between transgression and positive shifts in δ13C or regression and negative shifts is

complex (see time slices C1, C2, C4, C6 of their Fig. 3 where the coupling seems poor).

8

Conclusions and Implications

While I do not discount the presence of paleoceanographic effects on the isotope

record, based on the results seen here, I suggest that these effects are negligible in

certain circumstances. These Ordovician sections from Oklahoma and Maryland

demonstrate poor coupling between changes in the local sea-level and changes in δ13C.

Although there may have been local effects such as sea-water ‘ageing’ processes on at

least one part of our section at Clear Spring, Maryland, the δ13C signal observed appears

to be dominated by global changes in the carbon cycle. I postulate this based on the

correlation of the trend from different regional and tectonic settings and the poor

coupling between the shift and local sea-level. This positive shift in δ13C, if due to

increased organic carbon burial, may represent one of the first cooling events which led

to the glaciations of the Late Ordovician (Saltzman and Young, 2005).

Recommendations for Future Work

It would be useful for this study to have detailed results of the sequence

stratigraphy for the Middle-Upper Ordovician sections of the western United States for

additional comparison purposes. It would also be useful to have more δ13C data for this

time interval from other continents outside of Laurentia and Baltica to test whether the

positive shift seen is truly a global signal. Additionally, to further explore the question of

to what degree can changes in sea-level affect the carbon isotope record, it would be

interesting to model the timescale involved in δ13C fluctuations which are a result of sea-

level changes. For example, how long does it take for ‘ageing’ processes to become

significant? How many moles of organic carbon would need to be oxidized to produce a

+4‰ positive δ 13C shift? If high-amplitude flooding events do result in positive shifts

9

in δ13C, should this be recorded as an instantaneous change or can it be gradual? Or,

how much flooding is necessary to flush the platform and mix open marine waters with

the platform waters?

Acknowledgements

The author graciously thanks his advisor, Dr. Matthew Saltzman, without whom

this project would not be possible. Additionally, the work of Jesse Carlucci, Stephen

Westrop, Stephen Leslie, Seth Young, Stig Bergström, Alexa Sedlacek, Cole Edwards and

Charlie Diamond greatly contributed to this project with helpful discussion, sampling,

laboratory work, and other scientific contributions. I would like to thank the laboratory

team associated with the Stable Isotope Biogeochemistry Laboratory at the Ohio State

University (P.I. Andréa Grottoli, lab tech Yohei Matsui) and the biogeochemistry

laboratory team at Indiana University for conducting sample analyses. Funding for

analyses was provided by NSF EAR 0819832. The author also thanks the Friends of

Orton Hall for travel funds.

10

References

Ainsaar, L., Kaljo, D., Martma, T., Meidla, T., Männik, P., Nõlvak, J., and Tinn, O., 2010, Middle and Upper Ordovician carbon isotope chemostratigraphy in Baltoscandia: A correlation standard and clues to environmental history: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 294, no. 3–4, p. 189-201.

Allan, J., and Matthews, R., 1982, Isotope signatures associated with early meteoric diagenesis: Sedimentology, v. 29, no. 6, p. 797-817.

Bambach, K.R., Scotese, C.R., and Ziegler, A.M., 1980, Before Pangea: the geographies of the Paleozoic world: American Scientist, v. 68, p. 26-38.

Banner, J.L., and Hanson, G.N., 1990, Calculation of simultaneous isotopic and trace element variations during water-rock interaction with applications to carbonate diagenesis: Geochimica et Cosmochimica Acta, v. 54, no. 11, p. 3123-3137.

Blakey, R., 2008, Paleogeography and geologic evolution of North America. Retrieved from <http://www2.nau.edu/rcb7/namO450.jpg>

Brezinski, D.K., Taylor, J.F., and Repetski, J.E., 2012, Sequential development of platform to off-platform facies of the great American carbonate bank in the central Appalachians, in J.R. Derby, R.D. Fritz, S.A. Longacre, W.A. Morgan, and C.A. Sternback, eds., The great American carbonate bank: The geology and economic resources of the Cambrian-Ordovician Sauk megasequence of Laurentia: AAPG Memoir 98, p. 383-420.

Carlucci, J.R., 2012, Cyclicity and depositional sequences in a mixed siliciclastic –carbonate unit: Implications for intrabasinal correlation and subaqueous erosion in the Ordovician of Oklahoma: Geological Society of America Abstracts with Programs, v. 44, no. 7, p. 1222-1224.

Dickson, J., 2002, Fossil echinoderms as monitor of the Mg/Ca ratio of Phanerozoic oceans: Science, v. 298, no. 5596, p. 1222-1224.

Fanton, K., and Holmden, C., 2007, Sea-level forcing of carbon isotope excursions in epeiric seas: implications for chemostratigraphy: Canadian Journal of Earth Sciences, v. 44, p. 807-818.

Holmden, C., Creaser, R., Muehlenbachs, K., Leslie, S., and Bergström, S., 1998, Isotopic evidence for geochemical decoupling between ancient epeiric seas and bordering oceans: Implications for secular curves: Geology, v. 26, no. 6, p. 567-570.

Immenhauser, A., Kenter, J.A., Ganssen, G., Bahamonde, J.R., Van Vliet, A., and Saher, M.H., 2002, Origin and significance of isotope shifts in Pennsylvanian carbonates (Asturias, NW Spain): Journal of Sedimentary Research, v. 72, no. 1, p. 82-94.

11

Immenhauser, A., Della Porta, G., Kenter, J.A.M., and Bahamonde, J.R., 2003, An alternative model for positive shifts in shallow-marine carbonate δ 13C and δ 18O: Sedimentology, v. 50, no. 5, p. 953-959.

Johnston, D., Macdonald, F., Gill, B., Hoffman, P., and Schrag, D., 2012, Uncovering the Neoproterozoic carbon cycle: Nature, v. 483, no. 7389, p. 320-323.

Patterson, W.P., and Walter, L.M., 1994, Depletion of 13C in seawater ΣCO2 on modern carbonate platforms: Significance for the carbon isotopic record of carbonates: Geology, v. 22, no. 10, p. 885-888.

Payne, J.L., Lehrmann, D.J., Wei, J., Orchard, M.J., Schrag, D.P., and Knoll, A.H., 2004, Large perturbations of the carbon cycle during recovery from the end-Permian extinction: Science, v. 305, no. 5683, p. 506-509.

Read, J., 1980, Carbonate ramp-to-basin transitions and foreland basin evolution, Middle Ordovician, Virginia Appalachians: AAPG Bulletin, v. 64, no. 10, p. 1575-1612.

Saltzman, M.R., Runnegar, B., and Lohmann, K.C., 1998, Carbon isotope stratigraphy of Upper Cambrian (Steptoean Stage) sequences of the eastern Great Basin: Record of a global oceanographic event: Geological Society of America Bulletin, v. 110, no. 3, p. 285-297.

Saltzman, M.R., and Young, S.A., 2005, Long-lived glaciation in the Late Ordovician? Isotopic and sequence-stratigraphic evidence from western Laurentia.: Geology, v. 33, no. 2, p. 109-112.

Schmitz, B., Bergström, S.M., and Xiaofeng, W., 2010, The middle Darriwilian (Ordovician) δ13C excursion (MDICE) discovered in the Yangtze Platform succession in China: implications of its first recorded occurrences outside Baltoscandia: Journal of the Geological Society, v. 167, no. 2, p. 249-259.

Swart, P.K., 2008, Global synchronous changes in the carbon isotopic composition of carbonate sediments unrelated to changes in the global carbon cycle: Proceedings of the National Academy of Sciences, v. 105, no. 37, p. 13741-13745.

12

Arbuckle Mountains

Clear Spring

Fig. 1. Paleographic map of the Late Ordovician, approximately 450Ma. Credit: Ron Blakey (2008).

13

Fig. 2. Stratigraphic plots of 13

C from the Arbuckle Mountains, OK. Lithologies and sequence stratigraphy taken from Carlucci et al., (2012).

δ13

C

Arbuckle Mountains, OK

14

Fig. 3. Stratigraphic plot of 13

C from the Clear Spring, MD, sequence stratigraphy taken from Brezinski et al. (2012).

δ13

C

Clear Spring, MD

15

-9 -8 -7 -6 -5 -4

2

-3

1

0

-1

-2

-3

-4

δ13

C

δ18

O

Fig. 4. Cross plot of 13

C versus 18

O.

Clear Spring

Bromide

16

Fig. 5. Alternative model for explaining isotope shifts of δ13C and δ18O. Figure modified from Immenhauser et al., (2003).